Built by Metorial, the integration platform for agentic AI.
Translate text between 160+ languages with context awareness. Supports both single strings and batch translation of multiple strings in one request. The source language is auto-detected if not specified.
Generate concise summaries from text, URLs, or uploaded PDF files. Supports paragraph-style summaries or bullet-point lists. Provide exactly one content source: direct text, a URL to a PDF, or a file store key.
Validate text content for quality and safety issues. Combines spell checking, profanity detection, and spam detection into a single tool. Choose which checks to run. Returns detailed results for each enabled check including corrections, flagged words, and spam scores.
Generate vector embeddings for text, images, audio, or PDFs. Useful for semantic search, clustering, similarity comparison, and other vector-based operations. Provide the content and specify the type.
AI-powered web search that returns structured results with an optional AI-generated overview. Returns search results, image URLs, related links, and geographic data when relevant.
Intelligently scrape any website using AI to extract structured data. Provide a URL or raw HTML along with natural language prompts describing what data to extract, and the API returns structured results consistently across different sites. You can also use CSS selectors for precise targeting.
Forecast future values based on historical time series data. Provide a dataset of date-value pairs and get predicted future data points. Useful for trend analysis, demand forecasting, and financial projections.
Delete a file from JigsawStack's file storage by its key.
Generate images from text prompts using AI. Returns a URL to the generated image. Supports various aspect ratios, custom dimensions, and output formats (PNG, SVG). Advanced options include negative prompts, guidance scale, and seed for reproducibility.
Transcribe audio or video files to text using AI speech-to-text. Supports speaker diarization, language detection, and translation. Provide a URL to an audio/video file or a file store key. For long files, the result is returned directly; alternatively a webhook URL can be used for asynchronous processing.
Extract structured data from images or PDFs using vision OCR. Provide prompts describing what fields to extract (e.g., "total_price", "tax", "vendor_name" from a receipt). Supports JPEG, JPG, PNG, and multi-page PDFs. Returns extracted context, text sections with positioning data, and auto-generated tags.
Retrieve a previously uploaded file from JigsawStack's file storage. Returns the file's accessible URL. Files can be referenced by their key in other API calls (e.g., vOCR, speech-to-text).
Search for addresses, places, and points of interest. Optionally filter by country code or provide coordinates for proximity-based results. Also provides search suggestions for autocomplete functionality.
Analyze text to determine its emotional tone and sentiment. Returns an overall sentiment classification (positive, negative, or neutral) with a confidence score, plus a sentence-by-sentence breakdown with individual emotion and sentiment analysis.
Convert natural language queries into SQL statements. Provide a database schema for context and describe what you want to query in plain English. Supports PostgreSQL, MySQL, and SQLite dialects.
Detect not-safe-for-work content in images including nudity and gore. Returns boolean flags and confidence scores (0-1) for each category. Provide an image URL or file store key.
Identify and locate objects within images using AI. Returns detected objects with bounding boxes and labels. Optionally provide specific object prompts to search for, and request an annotated image with bounding boxes drawn.
Convert HTML content or a URL to images (PNG, JPEG, WebP) or PDF. Supports custom viewport sizes, full-page capture, dark mode, and various quality settings. Returns a URL to the generated file.